Theoretical studies of the reaction dynamics of the matrix-isolated F2+c i s-d2 -ethylene system

Abstract

The molecular dynamics of the F2+cis-d2 -ethylene addition reaction and the subsequent decomposition dynamics of the vibrationally excited 1,2-difluoroethane-d2 product isolated in Ar or Xe matrices at 12 K are investigated using trajectory methods that incorporate nonstatistical sampling to enhance the reaction probabilities. The matrix is represented by a face-centered-cubic crystal containing 125 unit cells with 666 lattice atoms in a cubic (5×5×5) arrangement. Both interstitial and substitutional sites for the F2/cis-d2 -ethylene pair are examined. Transport effects of the bulk are simulated using the velocity reset method introduced by Riley, Coltrin, and Diestler [J. Chem. Phys. 88, 5934 (1988)]. The potential-energy hypersurface for the system is written as the separable sum of a lattice potential, a lattice–substrate interaction, and a gas-phase potential for 1,2-difluoroethane-d2. The first two of these have pairwise form, while the 1,2-difluoroethane-d2 potential is identical to that employed previously to study the unimolecular reaction dynamics of matrix-isolated 1,2-difluoroethane-d4 [J. Chem. Phys. 93, 3160 (1990)]. The major F2+cis-d2 -ethylene reaction mechanism involves a four-center, concerted αβ addition across the C=C double bond. A small contribution from an atomic addition mechanism that initially forms fluoroethyl and fluorine radicals is observed in a xenon matrix, but not in argon. Subsequent to the formation of 1,2-difluoroethane-d2, the observed dynamic processes are vibrational relaxation to the lattice phonon modes, orientational exchange, and HF or DF elimination reactions. Vibrational relaxation is found to be very similar to that observed previously for 1,2-difluoroethane-d4. The process is well described by a first-order rate law with rate coefficients in the range 0.046–0.069 ps−1. The distribution of rate coefficients, as well as the averages, are nearly identical for Ar and Xe lattices. Very little difference is found between the relaxation rates for 1,2-difluoroethane-d2 and those for the HF(DF)+fluoroethylene products. The propensity for 1,2-difluoroethane-d2 to undergo orientational exchange increases as the available free space in the lattice decreases. Thus, it is a more important process in Ar than in Xe matrices. For the same reason, it occurs with greater frequency when the reactants are in an interstitial site than when they are substitutionally held. The probability of HF or DF elimination increases as the available free space in the matrix cage decreases. The relaxation rates show that this effect is not the result of different energy transfer rates. At least five distinct mechanisms play a role in HF and DF elimination reactions in the face-centered-cubic lattice. These are, in order of importance (a) αβ addition followed by syn elimination; (b) hydrogen- or deuterium-atom transfer to fluorine on the adjacent carbon followed by a protracted delay prior to C–F bond rupture; (c) rotation about the C=C double bond in the fluoroethylene product; (d) reversible hydrogen- or deuterium-atom transfer; and (e) atom addition with intervening delay. The computed elimination yield ratios between matrices are in good agreement with the experimental values. The calculated cis/trans ratio of fluoroethylenes formed subsequent to HF elimination in Ar are a factor of 2.7 lower than those observed in the experiments. The stabilization ratios are much larger than the experimental values. These results are interpreted to mean that the experimental matrix environment is more open and spacious than that for the crystal structure used in the calculations.